Project description by jury

The proposal describes an artificial intelligence-based method that enables the up-cycling of building components (e.g. concrete floor plates, glass elements, façade cladding) destined for dismantling. Starting from an indexed library of demolition rubble, a computational tool is able to guide the design of new buildings in order to optimize available materials. The entire library can be matched like jigsaw puzzle pieces onto adjustable shapes, producing significant money savings as well as a reduced carbon footprint for building construction. Material is placed in a holding pattern so that it maintains value and use through generations. The approach thereby attempts to create a less wasteful materials cycle by redeploying materials, and also reducing the volume of building rubble sent to landfill.

Jury appraisal

The LafargeHolcim Awards jury North America was very impressed by the pragmatism of the design method proposed in this research project that is able, in parallel, to promise a very compelling architecture. The idea of bringing new life to demolition rubble was found very relevant in terms of economic and environmental sustainability. The jury also appreciated the elegant and precise exemplifying design solutions that are featured along with the computational tool, which shows a clear sensitivity for architectural quality. In general, the idea of putting artificial intelligence combined with 3D printing at the service of both attractive design projects and resource conservation was found very powerful. The jury also considered the wide applicability in terms of building types and replicability in different geographic contexts a strong asset of the proposal.

AI and automation can transform the economics of down-cycling into up-cycling

Since the patent of the Bulldozer in 1932, buildings have been demolished without regard for the value of their material. There was public outcry in 1955 when demolition companies smashed up granite bathrooms in the Hotel Majestic West 72nd Street; the contractor argued “To recut the rosewood, mahogany and black walnut used in the interiors of the old glow room would have cost more than to use new materials." This project changes the economics of recycling building through automation. Conveyors of rubble are robotically scanned to form an indexed library. Machine learning algorithms trained to solve specific jigsaw puzzle problems are able to position each part into optimal configurations. Missing gaps can be custom fabricated, or robotic concrete printers extrude into targeted areas.

In the competition region of North America, the jury selected four entries to receive Next Generation prizes. The winning projects all address fundamental issues and challenges.